Introduction

Growth hormone (GH) exerts critical biological effects on the metabolic profile and cardiovascular (CV) system [1]. Congenital GH deficiency (GHD) occurs as isolated (IGHD) or combined with other pituitary hormone deficiencies (CPHD), with a prevalence of 1 in 3500–10,000 births [2]. Adult-onset GHD is associated with abnormal body composition, such as increased fat mass and reduced lean mass, impaired physical performance and an impact on the quality of life [3, 4]. Several studies have demonstrated an unfavorable CV risk in adults with untreated GHD, citing impaired glucose and lipid metabolism, arterial hypertension, endothelial dysfunction and arterial stiffness [5,6,7]. However, the impact of GH replacement therapy (GHRT) in adults with GHD has not been well established. Although the effects of GHRT on CV system have been demonstrated [8,9,10], it is not clear if the increased CV risk is due to GHD per se, or to confounding factors of acquired GHD (e.g. other pituitary hormone deficiencies with inadequate replacement, or consequences of pituitary surgery or radiotherapy) [11,12,13,14,15,16,17,18]. In adults with congenital GHD, the impact of GHD and its replacement on CV risk is poorly studied.

Cardiovascular diseases (CVD) are preceded by subclinical endothelial dysfunction, with increased arterial stiffness which results in accelerated atherosclerosis [19, 20]. In this context, carotid ultrasonography, carotid-femoral pulse wave velocity (cfPWV) and flow-mediated dilation (FMD) assessments are used in clinical practice as non-invasive, validated and reproducible techniques for subclinical atherosclerosis assessment [21], to stratify individuals according to CV risk.

However, to the best of our knowledge, no study has yet assessed metabolic profiles, body composition, carotid thickening, arterial stiffness, and endothelial function in adult patients with congenital GHD. Therefore, we evaluated these parameters in adult patients with congenital GHD, with and without GHRT.

Materials and methods

Study design

This cross-sectional single-center study was conducted at Hospital das Clinicas da Universidade de São Paulo, Brazil. Forty-nine adult patients with congenital GHD were included. The study was approved by the medical ethics committee of the Universidade de São Paulo. Written informed consent was obtained from all study subjects.

Subjects

From a congenital GHD cohort comprising 273 subjects, 49 adults with documented congenital GHD were invited to participate in this cross-sectional study, and all agreed to participate (Fig. 1). The inclusion criteria were: (a) proven adult GHD with congenital IGHD or CPHD; (b) outpatient clinic attendance and proven adherence to hormonal replacements for other pituitary deficiencies for at least 1 year; (c) for those with GHRT, IGF1 (insulin-like growth factor 1) levels between − 1 and 1 SD. Exclusion criteria were: (a) active smokers; (b) a diagnosis of diabetes mellitus or arterial hypertension; (c) taking medication known to interfere with glucose and lipid metabolism, blood pressure or the vascular system; and (d) patients with poor compliance.

Fig. 1
figure 1

Patient flow chart

Full size image

A GHD diagnosis was based on clinical, laboratory and imaging data. In childhood, short stature (Z-score <  − 2 or delta Z-score for target height <  − 1.5), low growth velocity, bone age delay, and a GH stimulatory peak after clonidine test of < 3.3 µg/L by immunofluorimetric assay [22]. Magnetic resonance imaging (MRI) of the hypothalamic-pituitary region was performed in all patients at the diagnosis of GHD to identify anatomic alterations. In the transition phase of the condition, which was characterized by the end of linear growth until complete adult maturation, patients were re-evaluated for GHD. In these cases, all patients with MRI abnormalities, low IGF1 values and GH peak < 5 µg/L in the stimulus test were considered as adults with congenital GHD [23].

Additional anterior pituitary hormonal deficiencies were also tested in individuals as recommended [24], and all pituitary hormonal deficiencies were under the appropriate replacement.

The adult subjects were divided into three groups:

  1. (1)

    Congenital GHD with GHRT during adulthood 29 patients (14 female), mean age = 36.6 ± 7.4 years, and with GHRT for 8 years (range 2–21 years). GH doses were approximately 1 IU/day (0.33 mcg/day) to maintain IGF1 in the normal range for age and sex. Thirteen patients used GHRT continuously (childhood and transition and adulthood). Sixteen patients used GHRT intermittently throughout life, and at the time of the study, all of them had been using GHRT for more than 2 years. Of these patients, 26 (90%) had CPHD and three (10%) had IGHD. Twenty-five patients (86%) with TSH deficiency had free T4 values (thyroxine) in the normal range (1.31 ± 0.4 ng/dl—reference range 0.9–1.7 ng/dl). Twenty-four patients (76%) had LH/FSH deficiency. Male patients had testosterone levels within the normal range (512 ± 196 ng/dL—reference range 249–740 ng/dL). For female patients, a regular menstrual cycle reflected good adherence to steroid hormone replacement. Of the 17 patients (56%) with ACTH deficiency, all were taking physiological doses of glucocorticoid therapy. The patients did not presented signs of supraphysiological doses. Seven were taking prednisone 5 mg/day, and 10 were taking hydrocortisone 15–20 mg/day.

  2. (2)

    Congenital GHD without GHRT during adulthood 20 patients (10 female), mean age = 37.4 ± 8.6 years, without GHRT in adulthood for 12 years (range 5–24 years). All patients never used GHRT in adulthood (16 used GHRT only in the childhood and/or transition phase and four never used GHRT during life). All subjects had been without GHRT for more than 5 years. Fourteen patients (70%) had CPHD and 6 (30%) had IGHD. Eleven patients (55%) with TSH (thyroid-stimulating hormone) deficiency had free T4 values in the normal ranges (1.2 ± 0.17 ng/dl—reference range 0.9–1.7 ng/dl). Thirteen patients (65%) had LH/FSH (luteinizing hormone/follicle-stimulating hormone) deficiency. All male patients had testosterone levels within normal ranges (485 ± 208 ng/dL—reference range 249–740 ng/dL). For female patients, menstrual cycles were regular, with good adherence to steroid hormone replacement. Seven patients (35%) with ACTH deficiency (adrenocorticotrophic hormone) were using glucocorticoid therapy at physiological doses. One was using prednisone at 5 mg/day, and six were using hydrocortisone at 15–20 mg/day. None showed clinical signs of supraphysiological glucocorticoid doses.

  3. (3)

    Control group 32 healthy individuals (17 females), mean age = 37 ± 8.9 years, all of whom had no evidence of cardiovascular disease after clinical and laboratory evaluations. Female controls do not use hormonal contraceptives.

Definition of premature CVD family history

Coronary artery disease in first-degree relative in male under 55 years old and/or female under 65 years old [25].

Definition of GHRT use

GHRT use was considered as follows: (1) For the groups of patients with GHRT during adulthood: the time was the number of years with uninterrupted GHRT; and (2) GHRT use throughout life: for all patients, the sum of years (not continuous) the patient used GHRT throughout their life.

Methods

Anthropometric and blood pressure measurements

Height and weight were measured, and body mass index (BMI) was calculated using the formula; BMI = weight (kg)/height2 (m). Abdominal waist (AW) (cm) was measured at the midpoint between the lower margin of the costal arches and the upper edge of the iliac crest. The waist-to-height ratio (WHR) was calculated using the abdominal waist (cm)/height (cm) formula [26]. Blood pressure (BP) was measured according to recommendations of the 7th edition of the Brazilian Hypertension guidelines [27].

Laboratory assessments

The total cholesterol (TC), high-density lipoprotein cholesterol (HDLc), low-density lipoprotein cholesterol (LDLc), triglycerides (Tg) levels were analyzed with an automatic enzymatic colorimetric method (Cobas Mira; F.Hoffmann-La Roche, Basel, Switzerland). The fasting glucose levels were determined with an automatic enzymatic method using hexokinase (Cobas Integra; Roche, Basel, Switzerland). Glycated hemoglobin (HbA1c) levels were measured by high-performance liquid chromatography (HPLC). The LH, FSH, total testosterone (TT), (17) estradiol and prolactin (PRL) levels were measured with immunofluorometric assays (Autodelfia, Turku, Finland) and more recently by electrochemiluminometric tests (Roche, Mannheim, Germany). The intra- and inter-assay coefficients of variation varied from 5 to 10%. Serum IGF1 levels were measured using a specific immunoradiometric assay (IRMA) or enzyme-labeled chemiluminescent immunometric assay (ICMA) and the values were transformed into SDS adjusted for sex and age (Siemens Healthcare Diagnostics, Dublin, Ireland). TSH and Free T4 were measured by electrochemiluminometric tests (Roche, Mannheim, Germany). For the control group, the following tests were performed: basal glucose, HbA1c, TC, HDLc, LDLc, Tg, IGF1, TSH and TT in men.

Assessment of carotid intima-media thickness

Carotid intima-media thickness (cIMT) measurements were performed by the same professional. A high-resolution ultrasound (GE, Vivid I, USA), with a high-frequency linear transducer, using B mode and a semi-automatic technique was used. A selected image was amplified to optimize visualization of the common carotid posterior wall, and the intima-media complex. The operator manually set measurement area points (start and end), and two lines along the artery were automatically drawn. Automated measurements included an online measure of multiple carotid intima-media data points (cm). The results were represented as the mean number of acquired data points.

Assessment of carotid-femoral pulse wave velocity

Carotid-femoral pulse wave velocity (cfPWV) (m/s) was assessed in participants using a Complior® device (Alam Medical, Vincennes, France). PWV was measured between the carotid and femoral artery using piezoelectric sensors; one was placed on the right side of the neck and the other on the femoral site. Sensor signals were recorded by the device software. The distance between sensors, as measured in a straight line, was used to approximate the arterial distance traveled by pulse waves. The foot of the pulse waves at both locations was used to calculate the mean cfPWV, once every five seconds [28]. This cfPWV measurement technique was previously described [29].

Assessment of flow-mediated dilation

The ultrasonography transducer was positioned on the brachial artery surface, on the right arm. For FMD (endothelium-dependent) measurements, a sphygmomanometer was inflated to a pressure of at least 50 mmHg above systolic pressure for 5 min, to induce a reactive hyperemia state. Images were captured by the transducer for 3 min after cuff release. For EIV (endothelium-independent vasodilatation) measurements, vasodilatation was evaluated after the administration of sublingual nitrate at dose of 5 mg. The artery diameter was evaluated using ultrasonography (Sequoia Echocardiography System® version 6.0, Siemens, CA, USA). Images were analyzed using an appropriate software package, managed by a professional blinded to group allocation. The basal phase and reactive hyperemia, pre- and post-nitrate diameter were analyzed using FMD Studio (Quipu, Pisa, Italy). This software provides basal diameter, pre-nitrate and post-nitrate diameter, and maximum arterial diameter values in the hyperemia phase. Women underwent the exam during the menstrual phase of their menstrual cycle because endothelial function in women during this period is similar to that observed in men at the same age [30, 31]. Four healthy subjects were also recruited to measure intra-observer reproducibility; they were evaluated at different times of the day by the same evaluator [28].

Assessment of total body densitometry using dual-energy X-ray absorptiometry

Whole-body densitometry was performed using dual-energy X-ray absorptiometry (DXA) on a Hologic DXA (Hologic, Inc. Crosby Drive, Bedford, MA, EUA) scanner to measure fat and lean mass of the whole body, except the head. The following parameters were recorded: (1) an index of total-body adiposity, derived by averaging the body fat percentage for soft tissue regions in spine and hip scans and (2) android/gynoid body aspect. The analysis was based on the following indices: (a) fat mass index (FMI) was calculated as fat mass (kg) divided by height (m) squared, considering normal values of 3–6 kg/m2 in men, and 5–9 kg/m2 in women [32], and (b) the Baumgartner index, which evaluated appendicular skeletal muscle mass, was calculated as the sum of skeletal muscle mass in the arms and legs (kg) divided by height (m) squared. According to Baumgartner’s anthropometric equation, sarcopenia was defined as < 5.5 kg/m2 for women, and < 7.26 kg/m2 for men [33, 34].

Statistical analyses

The Kolmogorov–Smirnov test was applied to test normal distribution. ANOVA or Student’s t-tests were used for parametrical data. ANOVA was followed by Bonferroni multiple comparisons if more than two variables showed significant difference. The Mann Whitney U test was used as a non-parametrical test. Fisher’s exact test and the χ2 test were used to estimate associations between categorical variables. Correlations were analyzed by Pearson's correlation coefficient.

A p value < 0.05 was considered statistically significant. The IBM-SPSS statistical package version 26.0 (Chicago, IL) was used for statistical analysis.

Results

General features

Congenital GHD with GHRT during adulthood

Twenty-nine patients, aged 36.6 ± 7.4 years, with an average height of 161.5 ± 13 cm (males; 168.8 ± 7.7 cm and females; 153.7 ± 13.4 cm) were included in this group. Five out of 29 (17%) had a familial history of CVD (Table 1). The median time of GHRT during childhood, the transition phase, adulthood and GHRT throughout life was 5.8 years (range 0–16 years), 3.1 years (range 0–7 years), 8 years (range 2–21 years) and 16.4 years (range 8.6–32 years), respectively.

Table 1 Clinical, anthropometric and laboratory characteristics of patients with and without GHRT, and controls
Full size table

TSH deficiency was treated with levothyroxine for a median time of 27 years (range 7–38 years). LH/FSH deficiencies were treated with sex steroids for a median time of 13.8 years (range 2–37 years). ACTH deficiency was treated with glucocorticoids for a median time of 19.3 years (range 1.4–31.5 years).

Congenital GHD without GHRT during adulthood

This group was made up of 20 patients aged 37.4 ± 8.6 years, with an average height of 155.8 ± 15.1 cm (males; 162.2 ± 14.2 cm and females; 149.4 ± 13.7 cm). Seven out of 20 (35%) had a familial history of CVD (Table 1). The median time without GHRT was 12 years (range 5–24 years). GHRT during childhood, the transition phase, and throughout life was 4 years (range 0–17 years), 0 years (range 0–7 years), and 7.6 years (range 0–19 years), respectively.

TSH deficiency was treated with levothyroxine for a median time of 22.5 years (range 13–44 years). LH/FSH deficiencies were treated with sex steroids for a median time of 15.4 years (range 6–38 years). ACTH deficiency was treated with glucocorticoids for a median time of 15.9 years (range 5.6–25.8 years).

Control group

The control group consisted of 32 healthy volunteers. Their mean age was 37 ± 8.9 years, and the mean height was 169 ± 0.1 cm (Table 1).

The patients currently undergoing GHRT had a longer GHRT time during the transition phase and throughout life than those without GHRT (p = 0.018 and p < 0.001, respectively) but not during childhood (p = 0.289). There was no statistical difference in the time of levothyroxine, sex steroids and glucocorticoids replacement between the groups (p = 0.860, p = 0.702 and p = 0.924, respectively).

Anthropometric, blood pressure, metabolic data and vascular properties

There were no significant differences between patients and controls regarding age, gender, CVD family history, AW measurements, systolic and diastolic blood pressure and glycemic profile. Height was statistically different between GHD patients without GHRT and controls (p value after Bonferroni multiple comparison = 0.002) (Table 1).

GHD patients with GHRT and controls had statistically lower WHR than patients without GHRT (p value after Bonferroni multiple comparison = 0.026 and < 0.001, respectively). Patients without GHRT had statistically higher overweight and obesity rates than patients with GHRT (40% vs.13.8%, respectively) (p = 0.048) (Table 1).

We observed statistically higher triglycerides and lower HDLc levels in patients without GHRT when compared with controls (p value after Bonferroni multiple comparison = 0.025 and = 0.029, respectively) but no statistical differences were found with GHRT group (Fig. 2).

Fig. 2
figure 2

Box plots of triglycerides and HDL cholesterol. a Triglyceride levels (mg/dL) in patients with GHRT vs. those without GHRT vs. controls (*p = 0.025). b HDLc levels (mg/dL) in patients with GHRT vs. those without GHRT vs. controls (*p = 0.029)

Full size image

No statistical differences were observed in large artery structural and functional vascular parameters between patients and controls (Table 2). However, a strongly and linear correlation existed between chronological age and cIMT (r = 0.645, p < 0.001) values in patients with congenital GHD, regardless GHRT (Fig. 3).

Table 2 Functional and structural arterial vessel features in patients with and without GHRT
Full size table
Fig. 3
figure 3

Positive associations between carotid intima media thickness (cIMT) for all adult patients with congenital GHD and age (r = 0.645, p < 0.001)

Full size image

Body composition by DXA

Fat percentages were lower in patients with GHRT, when compared to patients without GHRT (p = 0.003). Similarly, patients with GHRT had a significantly lower fat index, when compared to patients without GHRT (p = 0.041) (Table 3).

Table 3 Body composition in congenital GHD patients with and without GHRT (DXA analyses)
Full size table

We observed negative association between body fat percentage and FMI with GHRT throughout life (r =  − 0.378 p = 0.012 and r =  − 0.334 p = 0.029, respectively), which was independent of age. The Baumgartner index correlated inversely with age in patient with congenital GHD in adulthood (r =  − 0.388, p = 0.01) (Fig. 4).

Fig. 4
figure 4

a Negative association between body fat mass percentage in all adult patients with congenital GHD and GHRT throughout life (r =  − 0.378 p = 0.012). b Negative association between fat mass index (FMI) in all adult patients with congenital GHD and GHRT throughout life (r =  − 0.334 p = 0.029). c Negative associations between Baumgartner index in all adult patients with congenital GHD and age (r =  − 0.388 p = 0.01)

Full size image

Discussion

Several studies have identified associations between GHD and vasculometabolic impairments in adults [10]. Nevertheless, adult patients with congenital GHD have been under-represented in these studies. In the present study, we used strict exclusion criteria in congenital GHD adult patients to avoid confounding factors that could interfere with data interpretation. Thus, these patients are an interesting model to explore the impact of both GHD and GHRT on metabolic profiles and the CV system. There were neither structural nor functional large arterial differences in adult patients with congenital GHD, with or without GHRT, and healthy controls. However, the group without GHRT showed disturbances in lipid profiles and body composition reflected by WRH, BMI, fat mass percentages and fat mass index.

The impact of untreated GHD and GHRT on CV health remains controversial. A retrospective study of 1411 patients with untreated GHD observed an increase in general mortality, myocardial infarction and cerebrovascular events compared with the normal population [35]. This higher occurrence of CV events could be related to greater development of premature subclinical atherosclerosis evidenced by endothelial dysfunction [36], increased arterial stiffness and more atheromatous plaques in the carotid and femoral arteries in patients with GHD [37,38,39]. Low IGF-I levels were associated with increased cIMT, a recognized sign of premature subclinical atherosclerosis directly related to mortality from coronary artery disease [40]. Moreover, GHD is associated with increased levels of inflammatory cytokines, homocysteine and free radicals and reduced nitric oxide production [41]. Thus, these mechanisms appear to be important in endothelial homeostasis and vasodilation regulation.

Endothelial dysfunction occurs at the beginning of the atherosclerotic process, even before structural changes in vessel walls become evident. GHRT improves endothelial function by increasing flow-mediated dilation in patients with acquired GHD [10, 42,43,44]. A putative mechanism whereby GHRT improves vascular function occurs via IGF1-mediated stimulation of nitric oxide synthesis in endothelial cells [45]. A randomized, double-blind, placebo-controlled study observed the effect of GHRT over 6 months on arterial stiffness (assessed by PWV of the radial artery) and endothelial function (assessed by FMD of the brachial artery) in 32 adults with GHD, paired with controls of the same age and sex. The authors concluded that GHD in adults is associated with increased arterial stiffness and endothelial dysfunction and that GHRT improves these parameters. These findings suggest an important role for this treatment in reducing CV risk in these patients [10].

However, these beneficial results of the effect of GHRT on vascular properties have not been confirmed in all studies, and we did not observe them in our congenital patients. Van der Klaauw et al. [43] reported no change in cfPWV after 1.5 years of GHRT in 14 patients with acquired GHD. However, another study showed a significant reduction in cfPWV (8.1 to 6.7 m/s) during 6 months of GHRT in 16 patients with acquired GHD [46]. The discrepancies between these studies can be justified by the characteristics of the patients in the first study: older people, greater inclusion of men and higher BMI, factors related to a more disadvantageous CV profile.

A study in adults with congenital IGHD due to GHRHR gene mutation showed no evidence of premature atherosclerosis [47]. Furthermore, adults with congenital IGHD and never treated with GHRT have normal longevity [48]. In the same cohort, another study revealed increased CV risk, carotid thickening and atherosclerotic plaque development after GHRT with bimonthly depot of GHRT for 6 months [49]. Five years after GHRT withdrawal, carotid thickening had decreased to baseline values, but atherosclerotic plaque quantities did not change [50]. This study presents a notable contrast with most studies that had been published. Although we also found no changes in vascular properties, the effects of GHRT in IGHD were different to our patients, perhaps due to the application form of GHRT (daily versus depot) and the use of long-acting GHRT [51], which may have exposed body tissues to constant high GH levels [52].

Although GHD and its replacement had no impact on cIMT in our patients, aging was associated with increased carotid thickness, as observed in individuals without GHD. The influence of aging on vascular properties is well established [53].

Different degrees of GHD may result in different effects on the vascular wall [54]. We hypothesize that the pathophysiology of GHD in patients with congenital GHD is different from those with acquired GHD. The residual GH secretion may exert some metabolic actions, resulting in differences between these disease models. While a slight reduction in IGF1 in the general population (observed in patients with acquired GHD) has been associated with an increased risk of ischemic heart disease [55, 56], a more intense reduction in IGF1 (observed in patients with congenital GHD) can be protective against atherosclerosis [54]. These previous results are consistent with our findings.

The duration of the GHRT in GHD could be an important contributing factor to the results observed in this study. Patients with GHRT in adulthood were those who most replenished GH through their life, including during the transition phase, when GHRT improves changes in metabolic profile and body composition, reinforcing the importance of replacement during this period of life [57].

CfPWV and FMD are the gold standard methods to assess arterial stiffness and endothelial dysfunction in humans, respectively [21]. To the best of our knowledge, this is the first study to evaluate arterial stiffness and endothelial function in adult patients with congenital GHD. In our assistance, untreated GHD and the effects of GHRT in adults with congenital GHD did not accelerate subclinical atherosclerosis.

Modifications in lipid and glycemic profiles can also aggravate CV risk markers and contribute to CVD. Holdaway et al. [58], in their 3-year GHRT follow-up study, reported no significant changes in TC, LDLc, Tg, basal glucose and HbA1C levels. Fifteen years of GH replacement in GHD adults induced sustained improvements in serum lipid levels: TC and LDLc decreased and HDLc increased with no change in serum Tg level [9]. In our study, the group without GHRT showed higher Tg and lower HDLc levels compared with controls; the outcome displays a metabolic benefit of GHRT.

The literature data are inconsistent with regard to the glycemic profile, both in untreated and treated GHD patients. Adult patients with untreated GHD have increased visceral fat mass and often have an impaired glucose metabolism together with insulin resistance [59]. This association is based on studies with heterogeneous cohorts, different etiologies and severity of GHD. It is possible that the insulin resistance observed in these patients is due to the metabolic, inflammatory and body composition changes associated with GHD, but not directly caused by this condition. Castillo et al. [60], in their cross-sectional study, evaluated 15 patients with acquired GHD without GHRT, with adequate replacement of other pituitary hormone deficiency, and compared this cohort to a healthy group. They demonstrated that insulin sensitivity was similar to individuals with normal pituitary function, despite higher fat mass percentages in patients with GHD without GHRT. GHRT induces beneficial effects on body composition, findings that provide a rationale for improvement in insulin resistance with treatment. However, the glycemic effects of GHRT in GHD are conflicting. Whereas some studies have documented an improvement in glucose metabolism and insulin sensitivity [61], other investigations have observed no effect [62, 63].

In untreated congenital IGHD due to GHRHR gene mutation, insulin sensitivity is increased despite abdominal obesity [64]. Increased insulin sensitivity plays a fundamental role in longevity, possibly explaining the normal life span of these subjects [35]. As vascular properties, residual GH secretion can exert some metabolic actions, resulting in differences between these disease models (acquired vs. congenital). Visceral adiposity requires a minimal level of GH secretion to promote increased insulin resistance, which is not seen in patients with congenital GHD [64].

The absence of residual GH secretion and the adequate hormonal replacement of the other hormone deficient sectors of our patients, could explain the absence of glycemic alterations. These observations and controversies raise an important issue regarding insulin sensitivity and the role of GHD.

Lower GH levels are associated with an unfavorable impact on body composition as reported in acquired GHD patients [65]. In our study, there was a favorable effect of GHRT on WHR. The WHR metric is of great value because it is strongly associated with several chronic diseases, cardiovascular events and mortality rate [26].

The most consistent effect of GHRT in patients with GHD is a decrease in fat mass [9]. The impact of GHRT on body composition in adults with GHD was previously investigated in a randomized placebo-controlled trial (study duration range 2–18 months), which demonstrated a positive effect of GHRT on total lean and fat body mass, while BMI remained unaffected [6]. In our study, we identified lower BMI, WHR, fat percentage and FMI in the group with GHRT. Furthermore, prolonged use of GHRT was related to reduced fat body percentage. Our DXA approach evaluated fat and lean mass as a whole rather than compartmentalizing them; therefore, we were unable to differentiate total fat from visceral fat. With this limitation, we indirectly assessed visceral obesity using WHR. From the literature, WHR is the best predictor of whole-body fat percentages and visceral fat in men and women [66]. However, the ideal anthropometric parameter to predict CV risk in adults with congenital GHD is still unknown; thus, more studies must address whether WHR is a reliable marker of CV risk in these patients.

It is notable that GHRT did not impact lean mass in our study. Patients with untreated IGHD by GHRHR gene inactivation had better muscle strength parameters adjusted for weight and fat-free mass than controls and exhibited satisfactory muscle function [67]. In one open-label prospective study, patients with acquired GHD (61 men, mean age 50.0 years; range 22–74 years) were under GHRT for 10 years; they presented increased muscle strength during the first half of the study and were partially protected against normal age-related decline in strength during the last 5 years [68]. In a randomized, placebo-controlled crossover trial, 60 patients after more than 3 years of GHRT, a 4-month period of placebo treatment decreased the measured thigh muscle mass but without changing muscle strength [69]. Observed increases in lean mass are susceptible to measurement error and thus may occur without an improvement in patient strength [70].

In the current study, we only included patients with adequate adherence to hormonal treatment from other pituitary sectors. Inadequate pituitary hormone replacement therapies could also contribute to the CV risk. It has been shown that daily glucocorticoid doses higher than 20 mg of hydrocortisone are associated with an adverse metabolic profile in CPHD [71]. Furthermore, acquired CPHD patients with secondary adrenal insufficiency may frequently have residual cortisol secretion [72], which does not occur in congenital CPHD. In our cohort, all patients with ACTH deficiency were replaced with glucocorticoid at physiological doses. In addition, untreated hypogonadism might also exert an unfavorable impact on metabolism [73]. In our cohort, all patients with LH/FSH deficiency were receiving sex steroids. Only women with a regular menstrual cycle and men with normal testosterone levels were selected for the study. Finally, all patients were in replacement therapy with levothyroxine, with free T4 levels within the normal range. Adequate treatment of all pituitary hormonal sector deficits may have benefited from the absence of vasculometabolic changes in our patients. Our findings support further efforts to optimize replacement therapy for all deficiencies without compromising patient safety and well-being in the short and long term.

One strength of our study was the availability of data from a specifically well-selected sub-group of GHD patients. All patients had congenital GHD and were taking pituitary hormone replacements with adequate doses of glucocorticoids, levothyroxine and sex steroids. However, some limitations need to be considered. Due to the rarity of the congenital disease and the strict exclusion criteria adopted, our cohort was small. Another limitation is the cross-sectional nature of this study. Nevertheless, the design differs from many studies in the literature that have evaluated GHRT in vascular properties for small periods of therapy, without assessing the long-term effect on vasculature in GHD with and without GHRT. Given these limitations, prospective studies with larger patient samples and longer GHRT during adulthood are necessary to validate our results.

In conclusion, our study revealed no differences in structural and functional large arterial properties in congenital GHD patients with or without GHRT in adulthood. However, patients without GHRT had an unfavorable body composition. Adequate hormonal replacement of other pituitary sectors is also essential for vasculometabolic health. Furthermore, the data found in the current literature on acquired hypopituitarism are not directly applicable to the management of congenital GHD. These findings bring new insights into the treatment and follow-up of these patients and may guide GHD patients in their individual decision for GHRT in adulthood.